BioDecode Usage Guide

Everything you need to analyze your genome locally.

Contents

  1. System Requirements
  2. Installation & Activation
  3. First Launch & Database Setup
  4. Analyzing Your VCF File
  5. Understanding Your Report
  6. Using the Markdown Export with LLMs
  7. Updating the ClinVar Database
  8. FAQ

1. System Requirements

RequirementDetails
Operating SystemmacOS 12+ (Apple Silicon or Intel), Windows 10+, Linux
RAM4 GB minimum, 8 GB recommended
Disk Space~2 GB (ClinVar database + application)
InternetRequired only for initial ClinVar database download (~400 MB)

No Python or command line needed. BioDecode is a native desktop application. Everything is bundled — just download, install, and run.

2. Installation & Activation

Step 1: Download

After purchasing, you'll receive a license key by email. Download the BioDecode installer for your platform from the link in your email.

Step 2: Install

Step 3: Activate Your License

When you first open BioDecode, you'll see the activation screen. Paste the license key from your email and click Activate. Your license works offline — no internet connection required after activation.

One-time purchase, no subscription. Your license key never expires and works on any of your personal machines. No account creation needed.

3. First Launch & Database Setup

After activation, BioDecode will automatically set up the ClinVar database. This is a one-time process:

  1. Download — BioDecode downloads ~400 MB of ClinVar variant data from NCBI (the National Center for Biotechnology Information)
  2. Build — It processes 8.6 million rows and builds a local SQLite database with 4.28 million variants. This takes about 60-90 seconds.
  3. Ready — Once complete, you'll see the file drop zone and the green "ClinVar DB ready" indicator in the header.

You'll see a progress log showing every step in real time. No manual intervention needed.

All data stays on your machine. The ClinVar database is stored at ~/.biodecode/data/. Your genomic data is never uploaded anywhere.

4. Analyzing Your VCF File

Once the database is ready, you can analyze your genome:

  1. Drag and drop your .vcf or .vcf.gz file onto the drop zone, or click Browse Files to select it.
  2. Watch the progress — BioDecode parses your variants, matches them against ClinVar, and generates reports. For a typical whole genome (~4.7M variants), this takes about 30-60 seconds.
  3. View your report — The full HTML report appears directly in the app, with interactive tables linking to ClinVar and dbSNP.

BioDecode will:

Supported formats: VCF 4.x files, both uncompressed (.vcf) and BGZF-compressed (.vcf.gz). Optimized for GATK HaplotypeCaller output on the GRCh38 (hg38) reference genome.

5. Understanding Your Report

Your report appears inline in the app and is also saved to ~/.biodecode/output/ in three formats:

In-App Report

The full interactive report displays directly in BioDecode. It includes summary statistics, variant tables for each significance category, and clickable links to ClinVar and dbSNP. You can use Save as PDF to export it.

Saved Report Files

Significance Categories

CategoryMeaning
PathogenicVariant is classified as disease-causing in ClinVar
Likely PathogenicStrong evidence of being disease-causing, but not definitive
Uncertain Significance (VUS)Not enough evidence to classify as pathogenic or benign
Drug ResponseAffects how you metabolize or respond to specific medications
Risk FactorAssociated with increased risk for a condition
ConflictingDifferent labs have submitted conflicting classifications
Likely Benign / BenignVariant is not expected to cause disease

ClinVar Star Ratings

StarsMeaning
★★★★Practice guideline
★★★☆Reviewed by expert panel
★★☆☆Multiple submitters, no conflicts
★☆☆☆Single submitter
☆☆☆☆No assertion criteria provided

Important: Higher star ratings indicate stronger evidence. Variants with 0 stars should be interpreted with caution — they may be submitted without supporting evidence. Always prioritize findings with 2+ stars.

Zygosity

6. Using the Markdown Export with LLMs

BioDecode includes a de-identified Markdown report designed for sharing with AI language models to get plain-language explanations.

How to Use It

  1. After analysis, click Copy Markdown for LLM in the toolbar
  2. Open any LLM chat (ChatGPT, Claude, Gemini, etc.)
  3. Paste the copied content
  4. Ask for an explanation using one of these prompts:

Example Prompts

Privacy note: The Markdown report contains only variant-level data (gene names, positions, ClinVar classifications). No personally identifiable information is included. It's safe to share with any LLM service.

Remember: LLM-generated interpretations are NOT medical advice. AI models can make mistakes. The ClinVar data in your report is the authoritative source — the LLM is just helping you understand it in plain language. Always consult a genetic counselor for medical decisions.

7. Updating the ClinVar Database

ClinVar is updated weekly by NCBI. To get the latest variant classifications:

  1. Open Settings (gear icon in the top-right corner)
  2. Click Rebuild Database
  3. BioDecode will download the latest ClinVar data and rebuild the database automatically

This is free — ClinVar data is in the public domain (NCBI policy).

8. Frequently Asked Questions

What VCF formats are supported?

BioDecode supports VCF 4.x files, both uncompressed (.vcf) and BGZF-compressed (.vcf.gz). It's optimized for GATK HaplotypeCaller output on the GRCh38 (hg38) reference genome.

Does BioDecode work with GRCh37 (hg19)?

Currently, BioDecode matches against GRCh38 ClinVar coordinates only. If your VCF uses hg19, you'll need to lift over to hg38 first using a tool like UCSC LiftOver or Picard LiftoverVcf.

Does my data leave my computer?

No. BioDecode processes everything locally. The only network activity is the initial download of ClinVar data from NCBI's public FTP server. Your VCF file and reports never leave your machine.

Can I use BioDecode with 23andMe or AncestryDNA data?

Not directly. 23andMe and AncestryDNA provide genotyping chip data, not whole genome sequencing. Their raw data formats are different from VCF. You would need to convert their data to VCF format first. Tools exist for this, but coverage will be limited to the SNPs on their genotyping chip (~650K variants vs 4.7M+ for whole genome).

How should I interpret pathogenic variants?

A "pathogenic" classification in ClinVar means that variant has been reported as disease-causing by clinical laboratories. However, context matters:

What if I find something concerning?

Do not panic. Many people carry pathogenic variants without being affected, especially for recessive conditions (where you need two copies). The most important steps are:

  1. Note the gene, variant, and star rating
  2. Schedule an appointment with a genetic counselor
  3. They may recommend confirmatory testing through a CLIA-certified lab
  4. Do not make medical decisions based on BioDecode reports alone

Where are my reports saved?

Reports are saved to ~/.biodecode/output/ in HTML, JSON, and Markdown formats. Each analysis creates a timestamped set of files.

Can I run BioDecode on multiple computers?

Yes. Your license key works on any of your personal machines. Simply install BioDecode and enter your key.

Where can I get help?

Contact us at hello@biodecode.io for technical issues with the software.